SignificanceLower respiratory tract infections (LRTIs) are the leading cause of infectious disease-related deaths worldwide yet remain challenging to diagnose because of limitations in existing microbiologic tests. In critically ill patients, noninfectious respiratory syndromes that resemble LRTIs further complicate diagnosis and confound targeted treatment. To address this, we developed a metagenomic sequencing-based approach that simultaneously interrogates three core elements of acute airway infections: the pathogen, airway microbiome, and host response. We studied this approach in a prospective cohort of critically ill patients with acute respiratory failure and found that combining pathogen, microbiome, and host gene expression metrics achieved accurate LRTI diagnosis and identified etiologic pathogens in patients with clinically identified infections but otherwise negative testing.

Lower respiratory tract infections (LRTI) lead to more deaths each year than any other infectious disease category. Despite this, etiologic LRTI pathogens are infrequently identified due to limitations of existing microbiologic tests. In critically ill patients, non-infectious inflammatory syndromes resembling LRTI further complicate diagnosis. To address the need for improved LRTI diagnostics, we performed metagenomic next-generation sequencing (mNGS) on tracheal aspirates from 92 adults with acute respiratory failure and simultaneously assessed pathogens, the airway microbiome and the host transcriptome. To differentiate pathogens from respiratory commensals, we developed rules-based and logistic regression models (RBM, LRM) in a derivation cohort of 20 patients with LRTI or non-infectious acute respiratory illnesses. When tested in an independent validation cohort of 24 patients, both models achieved accuracies of 95.5%. We next developed pathogen, microbiome diversity, and host gene expression metrics to identify LRTI-positive patients and differentiate them from critically ill controls with non-infectious acute respiratory illnesses. When tested in the validation cohort, the pathogen metric performed with an AUC of 0.96 (95% CI = 0.86 - 1.00), the diversity metric with an AUC of 0.80 (95% CI = 0.63 – 0.98), and the host transcriptional classifier with an AUC of 0.88 (95% CI = 0.75 – 1.00). Combining these achieved a negative predictive value of 100%. This study suggests that a single streamlined protocol offering an integrated genomic portrait of pathogen, microbiome and host transcriptome may hold promise as a novel tool for LRTI diagnosis.SIGNIFICANCE STATEMENTLower respiratory tract infections (LRTI) are the leading cause of infectious disease-related death worldwide yet remain challenging to diagnose because of limitations in existing microbiologic tests. In critically ill patients, non-infectious respiratory syndromes that resemble LRTI further complicate diagnosis and confound targeted treatment. To address this, we developed a novel metagenomic sequencing-based approach that simultaneously interrogates three core elements of acute airway infections: the pathogen, airway microbiome and host response. We studied this approach in a prospective cohort of critically ill patients with acute respiratory failure and found that combining pathogen, microbiome and host gene expression metrics achieved accurate LRTI diagnosis and identified etiologic pathogens in patients with clinically identified infections but otherwise negative testing.

249/250 words)Lower respiratory tract infections (LRTI) lead to more deaths each year than any other infectious disease category(1). Despite this, etiologic LRTI pathogens are infrequently identified due to limitations of existing microbiologic tests(2). In critically ill patients, non-infectious inflammatory syndromes resembling LRTI further complicate diagnosis. To address the need for improved LRTI diagnostics, we performed metagenomic next-generation sequencing (mNGS) on tracheal aspirates from 92 adults with acute respiratory failure and simultaneously assessed pathogens, the lung microbiome and the host transcriptome. To differentiate pathogens from respiratory commensals, we developed rules-based and logistic regression models (RBM, LRM) in a derivation cohort of 20 patients with LRTI or non-infectious acute respiratory illnesses. When tested in an independent validation cohort of 24 patients, both models achieved accuracies of 95.5%. We next developed pathogen, microbiome diversity, and host gene expression metrics to identify LRTI-positive patients and differentiate them from critically ill controls with non-infectious acute respiratory illnesses. When tested in the validation cohort, the pathogen metric performed with an AUC of 0.96 (95% CI = 0.86 -1.00), the diversity metric with an AUC of 0.80 (95% CI = 0.63 -0.98), and the host transcriptional classifier with an AUC of 0.91 (95% CI = 0.80 -1.00).Combining all three achieved an AUC of 0.99 (95% CI = 0.97 -1.00) and negative predictive value of 100%. This study suggests that a single streamlined protocol offering an integrated genomic portrait of pathogen, microbiome and host transcriptome may hold promise as a novel tool for LRTI diagnosis.Page 4 of 41 SIGNIFICANCE STATEMENT (115/120 words) Lower respiratory tract infections (LRTI) are the leading cause of infectious disease-related death worldwide yet remain challenging to diagnose because of limitations in existing microbiologic tests. In critically ill patients, non-infectious respiratory syndromes that resemble LRTI further complicate diagnosis and confound targeted treatment. To address this, we developed a novel metagenomic sequencing-based approach that simultaneously interrogates three core elements of acute airway infections: the pathogen, lung microbiome and host response. We studied this approach in a prospective cohort of critically ill patients with acute respiratory failure and found that combining pathogen, microbiome and host gene expression metrics achieved accurate LRTI diagnosis and identified etiologic pathogens in patients with clinically identified infections but otherwise negative testing.

Accurate and informative microbiological testing is essential for guiding diagnosis and management of pneumonia in patients who are critically ill. Sampling of tracheal aspirate (TA) is less invasive compared with mini-bronchoalveolar lavage (mBAL) and is now recommended as a frontline diagnostic approach in patients who are mechanically ventilated, despite the historical belief that TA was suboptimal due to contamination from oral microbes. Advancements in metagenomic next-generation sequencing (mNGS) now permit assessment of airway microbiota without a need for culture and, as such, provide an opportunity to examine differences between mBAL and TA at a resolution previously unachievable. Here, we engaged shotgun mNGS to assess quantitatively the airway microbiome in matched mBAL and TA specimens from a prospective cohort of critically ill adults. We observed moderate differences between sample types across all subjects; however, we found significant compositional similarity in subjects with bacterial pneumonia, whose microbial communities were characterized by dominant pathogens. In contrast, in patients with noninfectious acute respiratory illnesses, significant differences were observed between sample types. Our findings suggest that TA sampling provides a similar assessment of airway microbiota as more invasive testing by mBAL in patients with pneumonia.